Quantile-Based Spectral Analysis of Time Series

Methods to determine, smooth and plot quantile periodograms for univariate and multivariate time series. See Kley (2016) for a description and tutorial.


quantspec 1.2-1

o Updated CITATION file as required by the JSS editor. o Updated vignette with editorial changes from JSS. o Updated references throughout documentation.

quantspec 1.2-0


o Added functionality for the analysis of multiple time series.


o Added plot function for LagOperator. o Fixed two typos in the computation of estimates for standard errors.

quantspec 1.1-0


o Added functionality to do lag window estimation.


o Fixed a wrong constant in the definition of the W1 kernel. o Fixed a tiny mistake in getLevels-QSpecQuantity.

quantspec 1.0-3

o Added spaces around = in some show commands. o Updated vignette to version that was accepted by the JSS. o Updated references in the technical documentation.

quantspec 1.0-2

o Reimplemented getSdNaive using Rcpp. o Moved the unit tests to tests/testthat, which is now recommended practice (cf. https://github.com/hadley/testthat). o Stylistically revised the vignette.

quantspec 1.0-1

o Fixed a mistake in QRegEstimator (in v1.0-0, the complex conjugate of the estimator was returned). o Fixed a mistake where 1 x 1 QuantileSD plots failed to be created. o Improved the vignette.

quantspec 1.0-0


o The quantspec package was conceptually remodeled. It is now based on an object-oriented design. The implementation uses the S4 system.


o efficient computation of frequency representations based on (1) the clipped time series, (2) the quantile regression estimator in the harmonic linear model.

o computation of the periodograms based on (1) and (2)

o computation of smoothed periodograms; using different types of weights will yield estimatiors for (1) the Laplace or copula spectral density (2) the integrated Laplace or copula spectral density

o simulation of (1) Laplace and copula spectral densities, (2) integrated Laplace and copula spectral densities,

o a mechanism to generate block bootstrap replicates in the time domain and compute all the estimators from them.


o A vignette was added to the package. It contains a brief introduction to quantile-based spectral analysis, a description of the framework, a tutorial and two worked examples.

quantspec 0.2 quantspec 0.1

o Three functions to compute, smooth and plot the (quantile regression-based) Laplace periodograms and rank-based Laplace periodograms.

Reference manual

It appears you don't have a PDF plugin for this browser. You can click here to download the reference manual.


1.2-3 by Tobias Kley, 2 years ago


Report a bug at http://github.com/tobiaskley/quantspec/issues

Browse source code at https://github.com/cran/quantspec

Authors: Tobias Kley [aut, cre] , Stefan Birr [ctb] (Contributions to lag window estimation)

Documentation:   PDF Manual  

Task views: Time Series Analysis

GPL (>= 2) license

Imports methods, graphics, quantreg, abind, zoo, snowfall, Rcpp

Depends on stats4

Suggests testthat

Linking to Rcpp

See at CRAN